Eight-probe Neuropixels recordings during spontaneous behaviors
datasetposted on 20.03.2019, 18:46 by Nick Steinmetz, Marius Pachitariu, Carsen Stringer, Matteo Carandini, Kenneth Harris
Datasets usually provide raw data for analysis. This raw data often comes in spreadsheet form, but can be any collection of data, on which analysis can be performed.
Eight-probe Neuropixels recordings in three mice of spontaneous activity.
Each "spks" is a structure of length 8, where each entry is a different probe (these probes were recorded simultaneously). It contains the spike times and the cluster identity of each spike.
load(fullfile(ephysroot, sprintf('spks%s_Feb18.mat' , mouse_name)));
spks(k).st = spike times (in seconds)
spks(k).clu = cluster identity of each spike in st (which neuron does spike belong to)
spks(k).Wheights = height of each cluster on the probe
Information about each probe's location is in "probeLocations.mat". The location of each site on the probe in microns in the Allen CCF framework is given in "ccfCoords". The brain area for each site is in "borders" as a function of the height of the site. To view the probe inside a wire-mesh brain, download and add to your path this toolbox: https://github.com/cortex-lab/allenCCF . You will also need to add to your path the npy-matlab toolbox: https://github.com/kwikteam/npy-matlab .
The behavioral file is the processed version of a mouse face movie (time x pixels x pixels). For raw videos, see this:
We processed these by taking the singular value decomposition of it.
beh = load(fullfile(ephysroot, sprintf('%s_face_proc.mat', mouse_name)));
beh.motionSVD = timepoints x components (left singular vectors * singular values)
beh.motionMask = pixels x pixels x components (right singular vectors)
beh.times = times of behavioral frames (in seconds) in same timeframe as spikes
See the "ephysLoad.m" script for how to load each into matlab and better understand the structures.